Oxygen Atom Transfer to Positive Ions: A Novel Reaction of Ozone in

This novel O-transfer reaction to positive ions, which expands our knowledge of the rich chemistry of ozone, introduces a new pathway for the gas-phas...
1 downloads 0 Views 108KB Size
J. Am. Chem. Soc. 1998, 120, 7869-7874

7869

Oxygen Atom Transfer to Positive Ions: A Novel Reaction of Ozone in the Gas Phase Maria Anita Mendes,† Luiz Alberto B. Moraes,† Regina Sparrapan,† Marcos N. Eberlin,*,† Risto Kostiainen,‡ and Tapio Kotiaho*,§ Contribution from the Institute of Chemistry, State UniVersity of Campinas, UNICAMP CP6154, 13083-970 Campinas, SP Brazil, Department of Pharmacy, DiVision of Pharmaceutical Chemistry, P.O. Box 56 (Viikinkaari 5), 00014 UniVersity of Helsinki, Helsinki, Finland, and VTT Chemical Technology, P.O. Box 140, FIN-02044 VTT, Espoo, Finland ReceiVed April 21, 1997

Abstract: In the gas phase, neutral ozone (O3) transfers an oxygen atom to several positive ions, i.e. the radical cations of pyridines (R-Py+•; R ) H, CH3, C2H5, and Cl), pyrimidine (Pi+•), and alkyl halides (CH3X+•; X ) Cl and I), and the halogen cations (X+; X ) Cl, Br, and I). Reactivity changes drastically within the halogen series (Cl+ , Br+ e I+), whereas no O-transfer occurs to F+. The oxide derivatives R-Py+-O•, Pi+-O•, CH3X+-O•, and XO+ are formed, as demonstrated by pentaquadrupole (QqQqQ) double- and triplestage mass spectrometry. No oxygen atom transfer occurs, however, in “inverse” reactions, i.e., those of ionized ozone (O3+•) with the corresponding neutrals; and charge transfer dominates. Ab initio calculations suggest that O-transfer from ozone to ionized pyridine yields ionized pyridine N-oxide Via simple nucleophilic addition of ozone as opposed to 1,3-dipolar cycloaddition. Similar nucleophilic addition followed by O2 loss is also the most likely mechanism for O-transfer from ozone to the ionized alkyl halides and halogen cations. This novel O-transfer reaction to positive ions, which expands our knowledge of the rich chemistry of ozone, introduces a new pathway for the gas-phase oxidation of halogen atoms, pyridines, pyrimidines, alkyl halides, and analogues, and consequently for the gas-phase generation of their chemically interesting but difficult to access ionized oxides.

Introduction Ozone (O3), the less stable form of elemental oxygen, plays a key role in the upper atmosphere by acting as a vital ultraviolet light screen.1 In solution, ozone acts as a strong oxidizing agent, an electrophile or nucleophile, and the diverse chemical behavior of ozone has been widely exploited in condensed-phase chemistry.2 The accumulation of electronegative oxygen atoms in the ozone molecule makes it also a very electrophilic 1,3dipole; hence it reacts promptly in concerted 1,3-dipolar cycloadditions with many olefins. Cycloaddition to olefins affords highly unstable trioxolanes, i.e. ozonides, the key intermediates in the cleavage of carbon-carbon double bonds by ozonolysis,3 an important reaction of ozone that is applied in synthetic and analytical procedures. Because of its vital importance in atmospheric and condensedphase chemistry, the physical and chemical properties of ozone in both its neutral and ionized forms have been studied extensively.1,4 The neutral/neutral and negative ion/neutral reactions of ozone have been studied thoroughly, but its positive ion/neutral reactions (O3+•/M or M+•/O3) have received much †

State University of Campinas. University of Helsinki. VTT Chemical Technology. (1) (a) McEwan, M. J.; Phillips, L. F. Chemistry of the Atmosphere; Edward Arnold Ltd., London, 1975. (b) Molina, M. J. Pure Appl. Chem. 1996, 68, 1749 (c) Slanger, T. G. Science 1994, 265, 1817. (2) (a) March, J. AdVanced Organic Chemistry, 3rd ed.; John Wiley & Sons: New York, 1985 (b) Carey, F. A.; Sunderberg, R. J. In AdVanced Organic Chemistry, 2nd ed.; Plenum Press: New York, 1984. (3) Bailey, P. S. Ozonation in Organic Chemistry; Academic Press: New York, 1978; Vol. 1. ‡ §

less attention. Few examples are reported, such as the reaction of neutral ozone with metal cations,1a and with strong Bronsted acids.5 The reactions of ozone with CH5+, for instance, allowed the first experimental detection of the elusive protonated ozone (O3H+), the estimation of its proton affinity, and the study of the reactions of O3H+ with methane.5 In the present study, ion/molecule reactions of neutral ozone with positive ions, i.e. ionized organic molecules (M+•) and halogen cations (X+), have been investigated in the gas phase by double- (MS2) and triple-stage (MS3)6 pentaquadrupole (QqQqQ) mass spectrometry,7 as well as those of ionized ozone with the corresponding neutrals (O3+•/M). Neutral ozone was observed to transfer readily an oxygen atom to several positive (4) (a) Smith, D.; Spanel, P. Mass Spectrom. ReV. 1995, 14, 255. (b) Atkinson, R.; Carter, W. P. L. Chem. ReV. 1984, 84, 437. (c) Steinfeld, J. I.; Adler-Golden, S. M.; Gallagher, J. W. J. Phys. Chem. Ref. Data 1987, 16, 911. (d) Wang, N. S.; Howard, C. J. J. Phys. Chem. 1990, 94, 8787. (e) Schmelz, T.; Chambaud, G.; Rosmus, P.; Ko¨ppel, H.; Cederbaum, L.; Werner, H.-J. Chem. Phys. Lett. 1991, 183, 209. (f) Domine´, F.; Ravishankara, A. R.; Howard, C. J. J. Phys. Chem. 1992, 96, 2171. (g) Viggiano, A. A. Mass Spectrom. ReV. 1993, 12, 115. (h) Hakola, H.; Arey, J.; Aschmann, S. M.; Atkinson, R. J. Atmos. Chem. 1994, 18, 75. (i) Fedchak, J. A.; Peko, B. L.; Champion, R. L. J. Chem. Phys. 1995, 103, 981. (j) Li, Z.; Friedl, R. R.; Sander, S. P. J. Phys. Chem. 1995, 99, 13445. (k) Arnold, S. T.; Morris, R. A.; Viggiano, A. A. J. Chem. Phys. 1995, 103, 2454. (l) Huey, L. G.; Hanson, D. R.; Howard, C. J. J. Phys. Chem. 1995, 99, 5001. (m) Newson, K. A.; Luc, S. M.; Price, S. D.; Mason, N. J. Int. J. Mass Spectrom. Ion Processes 1995, 148, 203. (n) Newson, K. A.; Price, S. D. Int. J. Mass Spectrom. Ion Processes 1996, 153, 151. (o) Garner, M. C.; Sherwood, C. R.; Hanold, K. A.; Continetti, R. E. Chem. Phys. Lett. 1996, 248, 20. (5) Cacace, F.; Speranza, M. Science 1994, 265, 208. (6) Schwartz, J. C.; Wade, A. P.; Enke, C. G.; Cooks, R. G. Anal. Chem. 1990, 62, 1809.

S0002-7863(97)01251-1 CCC: $15.00 © 1998 American Chemical Society Published on Web 07/28/1998

7870 J. Am. Chem. Soc., Vol. 120, No. 31, 1998

Mendes et al.

Figure 1. A diagram of the pentaquadrupole mass spectrometer. Q1, Q3, and Q5 are mass-analyzer quadrupoles whereas q2 and q4 function as ion-focusing reaction chambers. In a typical ion/molecule reaction experiment, ions are generated in the ion source, purified (mass-selected) by Q1, and further reacted under controlled conditions (collision energy and pressure) with a neutral gas introduced in q2. Product ions of interest are then mass selected by Q3, and structurally analyzed by collision-induced dissociation or structurally diagnostic ion/molecule reactions in q4, while Q5 is scanned to acquire the triple-stage mass spectra. For more details see ref 7f.

ions (M+•, X+) in a novel gas-phase oxidation reaction that affords the ionized N-oxide derivatives (M+-O•, XO+) as the main ionic products. Methods 2

3

6

The study was carried out via MS and MS experiments performed in an Extrel [Pittsburgh, PA] pentaquadrupole (QqQqQ) mass spectrometer, a versatile instrument for studying of ion/molecule reactions in the gas phase.7 In the QqQqQ8 (Figure 1) three mass analyzing (Q1, Q3, Q5) and two “rf-only” reaction quadrupoles (q2, q4) are placed sequentially. Ions are formed in the ion source, each desired ion is “purified” one at a time via mass selection in Q1, and their reactions are performed in q2 with selected neutrals at controlled conditions such as collision energy and concentration (neutral gas pressure). Product ions produced in q2 are mass selected one at a time by Q3 for further structural investigation in q4 via either structurally diagnostic collisioninduced dissociation (CID)9 or ion/molecule reactions, whereas Q5 is scanned to collect the triple-stage (MS3) product spectra. Collision energies of 1 eV (calculated by the m/z 39:41 ratio in neutral ethylene/ ionized ethylene reactions10) were applied for ion/molecule reactions in q2, whereas 15 eV CID with argon were performed in q4. The total pressures inside each differentially pumped region were typically 2 × 10-6 (ion-source), 8 × 10-6 (q2), and 8 × 10-5 (q4) Torr. (7) (a) Gozzo, F. C.; Eberlin, M. N. J. Am. Soc. Mass Spectrom. 1995, 6, 554. (b) Moraes, L. A. B.; Pimpim, R. S.; Eberlin, M. N. J. Org. Chem. 1996, 61, 8726. (c) Eberlin, M. N.; Sorrilha, A. E. P. M.; Gozzo, F. C.; Pimpim, R. S. J. Am. Chem. Soc. 1997, 119, 3550. (d) Moraes, L. A. B.; Gozzo, F. C.; Eberlin, M. N.; Vaniotalo, P. J. Org. Chem. 1997, 62, 5096. (e) Carvalho, M.; Moraes, L. A. B.; Kascheres, C.; Eberlin, M. N. J. Chem. Soc., Perkin Trans. 2 1997, 2347. (f) Eberlin, M. N. Mass Spectrom. ReV. 1997, 16, 113. (8) Juliano, V. F.; Gozzo, F. C.; Eberlin, M. N.; Kascheres, C.; Lago, C. L. Anal. Chem. 1996, 68, 1328. (9) (a) McLafferty, F. W. Tandem Mass Spectrometry; Wiley: New York, 1983. (b) Levsen, K.; Schwarz, H. Mass Spectrom. ReV. 1983, 2, 77. (c) Holmes, J. L. Org. Mass Spectrom. 1985, 20, 169. (d) Cooks, R. G.; Beynon, J. H.; Caprioli, C.; Lestes, J. R. Metastable Ions; Elsevier: New York, 1973. (e) Busch, K. L.; Glish, G. L.; McLuckey, S. A. Mass Spectrometry/Mass Spectrometry: Techniques and Applications of Tandem Mass Spectrometry; VHC: New York, 1988. (10) Tiernan, T. O.; Futrell, J. H. J. Phys. Chem. 1968, 72, 3080. (11) Newson, K. A.; Luc, S. M.; Price, S. D.; Mason, N. J. Int. J. Mass Spectrom. Ion Processes 1995, 148, 203. (12) Gaussian 94, Revision B.3, Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G. A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V. G.; Ortiz, J. V.; Foresman, J. B.; Peng, C. Y.; Ayala, P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian, Inc.: Pittsburgh, PA, 1995. (13) (a) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56, 2257. (b) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 72, 650. (14) Møller, C.; Plesset, M. S. Phys. ReV. 1934, 46, 618. (15) Dissociation occurs even under the low energy collision conditions used for the ion/molecule reactions likely owing to the high efficiency of transfer of translational to internal energy that occurs in quadrupole collision cells, see: Douglas, D. J. J. Am. Soc. Mass Spectrom. 1998, 9, 101. (16) Considering that very similar q2 collision conditions were applied for all reactions, the ratios of the sum of the abundance of ionic products to that of surving reactant ion were used as estimates of both reaction yields and relative reactivities.

Table 1. Ionic Products [m/z (Relative Abundance)a] of Reactions between Several Positive Ions and the O3/O2 Mixture

ion (M+• or X+)

O-transfer

charge transfer (O2+•)

pyridine+• b 3-chloropyridine+• b 2-methylpyridine+• 3-methylpyridine+• 4-methylpyridine+• pyrimidine+• 4-ethylpyridine+• pyrazole+• thiophene+• benzene+• bromobenzene+• benzyl bromide+• ethylene+• isoprene+• CH3CN+• CH3Cl+• b CH3I+• b NH2Cl+• c CH4+• F+ b Cl+ b 79Br+ b I+ b

95 (100) 129(100) 109(14) 109(27) 109(34) 96(100) 123 (42) none none none none none none none none 66(100) 158(100) none none none 51(3) 95(100) 143(100)

none none none none none none none none none none none none none none 32(100) 32(12) none none 32(100) 32(100) 32(100) none none

dissociation products of M+• 52(2), 39(2) 78(8) 65, 66, 67(3) 65, 66, 67(8) 65, 66, 67(9) 53(5) 106(12) 41(2) none 52(4), 51(1) 77(28) 91(7) 27(22) 67(8), 53(1) 40(6) 49(12) none none 15(19)

a When no signal of 100% relative abundance is reported, the base signal was the surving reactant ion. b Spectrum shown as a figure, see text. c Chloramine was sampled via membrane introduction mass spectrometry, see ref 19.

The ionized neutrals were generated by 70 eV electron ionization (EI) of their neutral counterparts, whereas the halogen cations were formed by dissociative 70 eV EI of CF3Cl, CH3Cl, CH3Br, and CH3I. Ozone was made from high-purity dried O2 in a custom-made ozonator and trapped on silica gel at -72 °C.11 The enriched O3/O2 mixture was introduced in q2 to obtain an instrument overall pressure of approximately 8 × 10-6 Torr (the actual pressure inside the collision cell is estimated to be close to 0.1 mTorr). Maximum yields of products were observed after 1-2 h of continuous flow of the gas mixture probably because of initial reactions of ozone with background compounds present in the gas lines. Ab initio molecular orbital calculations were carried out by Gaussian94.12 The species were optimized at the (HF) Hartree-Fock level of theory by employing the polarization 6-31G(d,p) basis set.13 Improved energies were obtained by single-point calculations at the HF/6-31G(d,p) level and incorporating valence electron correlation calculated by second-order Møller-Plesset (MP2) perturbation theory.14 Harmonic vibrational frequencies were calculated at the HF/6-31G(d,p) level and used to characterize the stationary points and to obtain the zero-point vibrational energies (ZPE).

Results and Discussion Several chemical species (mostly organic molecules), both in their neutral and ionized forms, were reacted with ionized or neutral ozone, respectively, via double-stage (MS2) experiments. The resulting product spectra display therefore the ionic products arising from collisions of the mass-selected ion with the neutral reactant.

O-Transfer from O3 to PositiVe Ions

J. Am. Chem. Soc., Vol. 120, No. 31, 1998 7871

Figure 2. Pentaquadrupole double-stage (MS2) product spectra for reactions of (a) ionized pyridine and (b) ionized 2-chloropyridine with neutral ozone. Note the O-transfer products of m/z 95 and 129.

M+•/O3 reactions. Most of the positively charged species react with neutral ozone (Table 1) mainly by charge transfer or undergo collision-induced dissociation,15 or both. However, several ionized pyridines, alkyl halides, and halogen atoms react with neutral ozone to yield mainly product ions of a formal O-transfer reaction, i.e. the ionic products display m/z ratios 16 units higher than that of the reactant ion. Ionized Pyridines. O-transfer from ozone to ionized pyridine occurs in high yield,16 and the product of m/z 95 represents more than 95% of the total product ion current (Figure 2a). The O-transfer product of ionized pyridine is best rationalized as ionized pyridine N-oxide (eq 1, R ) H, X ) CH). O-transfer from ozone also occurs in high to medium yields to several ionized pyridine derivatives, i.e. 2-, 3-, and 4-methylpyridine, 4-ethylpyridine (Table 1), 3-chloropyridine (Figure 2b), and the ionized 1,3-dinitrogen analogue pyrimidine (Table 1, eq 1, R ) H, X ) N).

Figure 3. Pentaquadrupole double-stage (MS2) product spectra for reactions of (a) F+, (b) 35Cl+, (c) 79Br+, (d) 81Br+, and (e) I+ with neutral ozone. Note that no O-transfer occurs to F+, whereas increasing yields of the XO+ product ions are observed in the following order: Cl+ , Br+ e I+.

(1)

Ionized Halogen Atoms. Figure 3 shows the product spectra for reactions of X+ (X ) F, Cl, Br, I) with neutral ozone. O-transfer to X+ is not general, and a drastic change in reactivity occurs. Note in Figure 3 that the higher the atomic number of the halogen atom the higher the yield16 of the O-transfer product XO+. In fact, no product of O-transfer to F+, i.e. FO+ of m/z 35, is formed (Figure 3a), and O2+• of m/z 32 dominates. Several pathways may account for formation of O2+• in the F+/ O3(O2) reactions. F+ may undergo charge transfer with O3, and O3+• may dissociate subsequently to O2+• or undergo charge transfer with O2 present in the reaction mixture. Charge transfer may, however, occur directly from F+ to O2. Charge transfer of F+ to both O3 and O2 is expected to be highly favored owing to the high ionization potential (IP) of the F atom (17.42 eV) when compared to that of O3 (12.43 eV) and O2 (12.07 eV).17 O-transfer may in fact occur from ozone to F+, but FO+ could be consumed rapidly owing to favorable charge transfer to O3 or O2 [the FO• radical (12.77 eV) has a slightly greater ionization (17) Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. D.; Mallard, W. G. J. Phys. Chem. Ref. Data 1988, 17, Suppl. 1.

energy than O2 (12.07 eV) and O3 (12.43 eV)17]. To investigate this possibility, reactions of F+ with the O3/O2 mixture were performed under mainly single collision conditions by using much lower pressures in q2. The secondary reaction of FO+ with O3 or O2 should be disfavored under single collision conditions; hence, if indeed FO+ is formed as the primary product, it should be detected in the low-pressure product spectra. No FO+ was detected, however, when working on a range of low-pressure conditions. This finding suggests that if O2+• is formed by charge transfer, it must occur directly from the primary F+ reactant ion. An alternative remains for a process suppressing O-transfer from O3 to F+, i.e. the highly electrophilic F+ may abstract O-• from ozone (oxygen anion transfer), a reaction that would yield O2+• and the FO• radical. The IP of the chlorine radical (12.97 eV) is just slightly higher than that of O3 and O2; hence reactions of Cl+ with the O3/O2 mixture occur predominantly by charge transfer or alternatively by oxygen anion transfer (O2+• of m/z 32), but also modestly by O-transfer yielding ClO+ of m/z 51 (Figure 3b). Note that secondary charge transfer of ClO+ to O3/O2 is not expected because the ionization energy of the ClO• radical (10.95 eV)17 is considerably lower than that of O3 and O2.

7872 J. Am. Chem. Soc., Vol. 120, No. 31, 1998 Table 2.

Mendes et al.

Ionic Products [m/z (Relative Abundance)] of Reactions between Ionized Ozone (O3+•) and Several Neutral Compounds

neutral reactant (M)

charge transfer (M+•)

O2+•

M+• dissociation products

other products

CH2dCH2 CH2dCH-CH2-Cl isoprene benzene styrene furan thiophene pyridine pyrimidine anisole 3-chlorothiophene CH3-S-CH3 C2H5-S-S-C2H5 CH3I CH3Cl CO CO2

28(100) 76, 78(100) 68(10) 78(100) 104(100) 68(100) 84(100) 79(47) 80(100) 108(100) 118, 120(100) 62(100) 122(100) 142(100) 50(100) none none

32(58) 32(20) 32(18) 32(8) 32(4) 32(3) 32(3) 32(68) 32(45) 32(34) 32(5) 32(15) 32(28) 32(1) 32(7) 32(100) 32(100)

none 41(57) 67(100) none none none none none none none none 61(98), 49(8) 94, 66(3) none 49(25) none none

41(100), 69(18)a 81(57) 81, 95, 107, 135 (23)b none none none none 80(100)c 81(82)c none none none none 157(32)d none none none

a Secondary products of ionized ethylene. b Secondary products of ionized isoprene, see ref 20. c Protonated Neutral. d The secondary CH3I+-CH3 product.

Figure 4. Pentaquadrupole double-stage (MS2) product spectra for reactions of (a) CH3I+• and (b) CH3Cl+• with neutral ozone. Note the O-transfer product of m/z 158 and 66.

The Br (11.81 eV) and I (10.45 eV) atoms display IP’s considerably lower than that of O3; hence, in reactions of Br+ (Figure 3c,d) and I+ (Figure 3e) with O3, the O-transfer products 79BrO+ of m/z 95 and IO+ of m/z 143 are formed in high yields,16 i.e. they constitute practically 100% of the product ion current. O-transfer from O3 to 79Br+ (Figure 3c) was confirmed by reacting 81Br+ (Figure 3d) and noting formation of 81BrO+ of m/z 97. Ionized Alkyl Halides. O-transfer is also a major reaction of ozone with ionized methyl iodide (Figure 4a) and methyl chloride (Figure 4b). Simple mechanistic considerations suggest the corresponding oxides CH3X+-O• (X ) Cl, I) to be the most likely products (eq 2). (2)

“Inverse” M/O3+• Reactions. Ionized ozone (O3+•) was also reacted with some neutrals, including pyridine, pyrimidine, methyl chloride, and methyl iodide (Table 2). However, no O-transfer products were formed, and charge transfer dominates. Both the lack of O-transfer and the dominance of charge transfer for the M/O3+• reactions are rationalized in terms of the lower

IP’s of the neutrals (for instance, IP of pyridine ) 9.25 eV) when compared to that of ozone (12.43 eV).17 After charge transfer, however, the ion/neutral complexes in the “inverse” reactions are the same as those in the “direct” reactions [M + O3+• f M+•/O3 r M+• + O3]. An obvious difference between the ion/neutral complexes from both reactions is their internal-energy content. Therefore, even the “inverse” reaction could become efficient in promoting Otransfer if the energy content of the ion/neutral complex is reduced. Quenching of the ion/neutral complex could be achieved, for instance, by varying the translational energy of the O3+• reactant or via collisional deactivation. Thus, the reactions of O3+• with all the neutrals reported in Table 2 were performed in the whole range of translational energies in which reactions occur, i.e. from nominal -1 eV to 3 eV.10 In another effort to promote collisional quenching of the ion/neutral complexes, the neutral gas pressures in the collision cell (q2) were varied to cause depletion of the O3+• current from 5% to 95%. No O-transfer products were detected in any of these conditions. Hence, either the nascent neutral ozone is much too “hot” and dissociates rapidly after exothermic charge transfer (for instance reaction of O3+• with neutral pyridine gives 3.18 eV17 excess energy after charge transfer) or the ion/neutral complexes M+•/O3 formed in the “inverse” reactions are too short-lived for collisional quenching. The Neutral Reactant: O3 or O2? Owing to incomplete conversion of O2 to O3 and possible degradation of O3 in the q2 gas lines, a considerable amount of O2 is present in the reaction mixture.11 Thus, O2 could participate in the O-transfer reaction. To verify this possibility, the ions for which O-transfer occurs were reacted with pure O2 under the same experimental conditions applied to the reactions with the O3/O2 mixture. Conditions were adjusted to maximize O-transfer with the O3/ O2 mixture; then this mixture was replaced by pure O2. No O-transfer products remained, which confirms that O-transfer occurs from O3 alone. The Structures of the O-Transfer Products. The multiple ion-selection/reaction capabilities of the pentaquadrupole mass spectrometer8 (Figure 1) permitted “on-line” access to structural information of the O-transfer products via the recording of their (18) Stevenson, D. P. Discuss. Faraday Soc. 1951, 10, 35. (19) (a) Kotiaho, T.; Eberlin, M. N.; Shay, B. J.; Cooks, R. G. J. Am. Chem. Soc. 1993, 115, 1004. (b) Eberlin, M. N.; Kotiaho, T.; Cooks, R. G. J. Am. Chem. Soc. 1994, 116, 2457. (20) Eberlin, M. N.; Cooks, R. G. J. Am. Chem. Soc. 1993, 115, 9226.

O-Transfer from O3 to PositiVe Ions

J. Am. Chem. Soc., Vol. 120, No. 31, 1998 7873

Figure 5. Pentaquadrupole (a) double-stage (MS2) dissociation product spectrum of ionized pyridine N-oxide generated by 70 eV EI of the neutral and (b) triple-stage (MS3) sequential product spectra of the O-transfer products arising from reactions of neutral ozone with ionized pyridine. The similarity of both spectra suggests O-transfer to the pyridine nitrogen.

sequential product spectra,6 typical examples of which are presented in Figures 5 and 6. To acquire these triple-stage (MS3) mass spectra, the O-transfer products were mass selected by Q3 and then subject to 15 eV dissociative collisions with argon in q4, whereas Q5 was scanned across the appropriate m/z range. The triple-stage mass spectrum of the product of O-transfer to ionized pyridine is shown in Figure 5b, whereas Figure 5a shows the double-stage spectrum of an authentic ion, i.e. ionized pyridine N-oxide formed by direct ionization of the corresponding neutral. Both spectra are nearly identical, which provides solid evidence that Py+-O• is formed as the primary O-transfer product. 79BrO+ (Figure 6a) and IO+ (Figure 6b) dissociate exclusively to 79Br+ of m/z 79 and I+ of m/z 127, respectively, as predicted by the Stevenson’s rule.18 This classic rule of ion dissociation states that the fragment ion of lower IP is formed preferentially. Hence, Br+ and I+ are the expected fragments considering the following IP’s: O (13.62 eV), Br• (11.81), and I• (10.45).17 CH3I+-O• of m/z 158 (Figure 6c) dissociates upon collision activation mainly by CH3• loss to afford IO+ of m/z 143 (Figure 3d), whereas IO+ apparently dissociates in turn to I+ of m/z 127. In contrast, CH3Cl+-O• of m/z 66 dissociates predomiTable 3.

nantly by ClO• loss (Figure 6d) to yield CH3+ of m/z 15, as expected from the lower IP of the methyl radical [ClO• (10.95 eV) and CH3• (9.84 eV)]. Ab Initio Calculations. Mechanism of O-Transfer to Pyridine. O-transfer of ozone to ionized pyridine could proceed via (a) a simple nucleophilic addition followed by O2 loss to afford ionized pyridine N-oxide or (b) 1,3-dipolar cycloaddition forming primarily an 1,2,3-trioxolane bicyclic intermediate (Scheme 1). Direct dissociation by O2 loss of the bicyclic intermediate would also yield ionized pyridine N-oxide, whereas prior isomerization via ring expansion to the corresponding 1,2,4-trioxolane bicyclic ring system (a process normally

Total Energies from Structure Optimization MP2/6-31G(d,p)//6-31G(d,p) +ZPE ab Initio Calculations

a

a

Figure 6. Pentaquadrupole triple-stage (MS3) sequential product spectra of the oxygen atom transfer products (a) 79BrO+, (b) IO+ (c) CH3IO+•, and (d) CH3ClO+•.

b

d

c

f

e

species

MP2/6-31G(d,p)//6-31G(d,p) energy (hartree)

ZPE (hartree)a

MP2/6-31G(d,p)//6-31G(d,p)+ZPEa energy (hartree)

a O3b O2b b c d e f

-247.19081 -224.79448 -149.94681 -472.10457 -472.06540 -472.05317 -322.12024 -322.14345

0.08201 0.00603 0.00405 0.09550 0.09538 0.09697 0.08280 0.08552

-247.10880 -224.78844 -149.94276 -472.00908 -471.97002 -471.95621 -322.03744 -322.05793

ZPE energies were scaled by 0.89. b Energies of the most stable triplete state.

7874 J. Am. Chem. Soc., Vol. 120, No. 31, 1998

Mendes et al.

Scheme 1

postulated for ozonolysis3) could produce an isomeric O-transfer product, i.e. ionized 1,2-oxazepine (Scheme 1). To compare the energetics of these two alternative reaction pathways, an ab initio potential energy surface diagram (Figure 7) was elaborated (Table 3). Simple nucleophilic addition of ozone to ionized pyridine is shown to be far more exothermic (-70.1 kcal/mol) than cycloaddition (-45.7 kcal/mol). The less favorable 1,2,3-trioxolane cycloadduct, if formed to any extent, is predicted to dissociate directly by O2 loss to yield also ionized pyridine N-oxide in a -6.3 kcal/mol exothermic process. Although the alternative isomerization of the 1,2,3-trioxolane to the 1,2,4-trioxolane cycloadduct, and further O2 dissociation, would generate the more stable O-transfer product, i.e. ionized 1,2-oxazepine (Figure 7), both the endothermicity of the initial isomerization (8.7 kcal/mol) and the substantial high energy barrier expected for the isomerization process must suppress this reaction pathway. Therefore, from the calculations summarized in Figure 7, O-transfer from ozone to pyridine proceeds mainly via simple nucleophilic addition of ozone and generates, upon O2 loss, ionized pyridine N-oxide as the main ionic product. This theoretical prediction agrees with the MS3 experiments that, as already discussed, also suggest formation of ionized pyridine N-oxide. Note that isomerization of ionized pyridine N-oxide to the more stable 1,2-oxazepine isomer is expected to be hampered by a substantial energy barrier for ring expansion. This isomerization may occur to some extent, however, under collisional activation, as the HCN loss that occurs for both the O-transfer product of pyridine and the authentic ionized pyridine N-oxide (the fragment of m/z 68 in Figure 5) is best rationalized from the isomeric 1,2-oxazepine structure. Concluding Remarks Oxygen anion (O-•) and O-transfer are commonly observed in negative ion-neutral reactions and neutral-neutral reactions

Figure 7. Ab initio MP2/6-31G(d,p)//6-31G(d,p) + ZPE potential energy surface diagram for reactions of ionized pyridine with neutral ozone. Energy barriers were not estimated. Energies are given in kcal/ mol. The less energetic reaction pathway is represented by simple nucleophilic addition followed by O2 loss that yields ionized pyridine N-oxide.

of ozone.4 For example, it has been observed that neutral ozone transfers an oxygen atom to CH3S•, CH3SS•,4f and HS• 4d and to NO2-.4c It has also been reported that O3-• transfers an oxygen anion (O-•) to CO24a and NO2.4k Therefore, the reactivity of O3+• described herein contrasts to that of its negative counterpart O3•-, i.e. O3+• reacts with a variety of neutral species exclusively by charge transfer whereas O-transfer is common for O•- The present results show, however, that O-transfer is achieved when the reaction is performed with neutral ozone and some selected positively charged species. In the 70’s it was reported that O-transfer occurs from O3 to inorganic ions such as NO+, K+, Fe+, Mg+, and Cu+.1a For the first time, therefore, neutral ozone is shown to transfer efficiently an oxygen atom to halogen cations, most efficiently to Br+ and I+, and two classes of organic ions, i.e. ionized alkyl halides and ionized nitrogenated aromatic compounds. This novel O-transfer reaction, which reveals a new pathway by which ozone acts as a strong oxidizing agent, opens a new route to generate and study several ionized oxides in the gas phase. They include ionized pyridine and pyrimidine N-oxides, and some difficult to access but chemically fascinating species such as CH3X+-O• and the oxohalogen cations XO+. Acknowledgment. This work was supported by the Research Foundation of the State of Sa˜o Paulo (FAPESP) and the Brazilian National Research Council (CNPq). T.K. acknowledges FAPESP for a visiting researcher scholarship. JA971251J